edsl 0.1.37.dev3__tar.gz → 0.1.37.dev4__tar.gz

{edsl-0.1.37.dev3 → edsl-0.1.37.dev4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: edsl
-Version: 0.1.37.dev3
+Version: 0.1.37.dev4
 Summary: Create and analyze LLM-based surveys
 Home-page: https://www.expectedparrot.com/
 License: MIT

edsl-0.1.37.dev4/edsl/__version__.py

@@ -0,0 +1 @@
+__version__ = "0.1.37.dev4"

{edsl-0.1.37.dev3 → edsl-0.1.37.dev4}/edsl/coop/coop.py

@@ -568,7 +568,9 @@ class Coop:
             "version": data.get("version"),
         }
 
-    def remote_inference_cost(self, input: Union[Jobs, Survey]) -> int:
+    def remote_inference_cost(
+        self, input: Union[Jobs, Survey], iterations: int = 1
+    ) -> int:
         """
         Get the cost of a remote inference job.
 
@@ -593,6 +595,7 @@ class Coop:
                     job.to_dict(),
                     default=self._json_handle_none,
                 ),
+                "iterations": iterations,
             },
         )
         self._resolve_server_response(response)

{edsl-0.1.37.dev3 → edsl-0.1.37.dev4}/edsl/inference_services/AwsBedrock.py

@@ -28,12 +28,16 @@ class AwsBedrockService(InferenceServiceABC):
         "ai21.j2-ultra",
         "ai21.j2-ultra-v1",
     ]
+    _models_list_cache: List[str] = []
 
     @classmethod
     def available(cls):
         """Fetch available models from AWS Bedrock."""
+
+        region = os.getenv("AWS_REGION", "us-east-1")
+
         if not cls._models_list_cache:
-            client = boto3.client("bedrock", region_name="us-west-2")
+            client = boto3.client("bedrock", region_name=region)
             all_models_ids = [
                 x["modelId"] for x in client.list_foundation_models()["modelSummaries"]
             ]
@@ -80,7 +84,8 @@ class AwsBedrockService(InferenceServiceABC):
                     self.api_token
                 )  # call to check the if env variables are set.
 
-                client = boto3.client("bedrock-runtime", region_name="us-west-2")
+                region = os.getenv("AWS_REGION", "us-east-1")
+                client = boto3.client("bedrock-runtime", region_name=region)
 
                 conversation = [
                     {

{edsl-0.1.37.dev3 → edsl-0.1.37.dev4}/edsl/jobs/Jobs.py

@@ -193,7 +193,7 @@ class Jobs(Base):
                     inference_service=invigilator.model._inference_service_,
                     model=invigilator.model.model,
                 )
-                costs.append(prompt_cost["cost"])
+                costs.append(prompt_cost["cost_usd"])
 
         d = Dataset(
             [
@@ -241,10 +241,25 @@ class Jobs(Base):
 
         try:
             relevant_prices = price_lookup[key]
-            output_price_per_token = 1 / float(
-                relevant_prices["output"]["one_usd_buys"]
+
+            service_input_token_price = float(
+                relevant_prices["input"]["service_stated_token_price"]
+            )
+            service_input_token_qty = float(
+                relevant_prices["input"]["service_stated_token_qty"]
+            )
+            input_price_per_token = service_input_token_price / service_input_token_qty
+
+            service_output_token_price = float(
+                relevant_prices["output"]["service_stated_token_price"]
             )
-            input_price_per_token = 1 / float(relevant_prices["input"]["one_usd_buys"])
+            service_output_token_qty = float(
+                relevant_prices["output"]["service_stated_token_qty"]
+            )
+            output_price_per_token = (
+                service_output_token_price / service_output_token_qty
+            )
+
         except KeyError:
             # A KeyError is likely to occur if we cannot retrieve prices (the price_lookup dict is empty)
             # Use a sensible default
@@ -254,9 +269,8 @@ class Jobs(Base):
             warnings.warn(
                 "Price data could not be retrieved. Using default estimates for input and output token prices. Input: $0.15 / 1M tokens; Output: $0.60 / 1M tokens"
             )
-
-            output_price_per_token = 0.00000015  # $0.15 / 1M tokens
-            input_price_per_token = 0.00000060  # $0.60 / 1M tokens
+            input_price_per_token = 0.00000015  # $0.15 / 1M tokens
+            output_price_per_token = 0.00000060  # $0.60 / 1M tokens
 
         # Compute the number of characters (double if the question involves piping)
         user_prompt_chars = len(str(user_prompt)) * get_piping_multiplier(
@@ -279,7 +293,7 @@ class Jobs(Base):
         return {
             "input_tokens": input_tokens,
             "output_tokens": output_tokens,
-            "cost": cost,
+            "cost_usd": cost,
         }
 
     def estimate_job_cost_from_external_prices(
@@ -326,7 +340,7 @@ class Jobs(Base):
                         "system_prompt": system_prompt,
                         "estimated_input_tokens": prompt_cost["input_tokens"],
                         "estimated_output_tokens": prompt_cost["output_tokens"],
-                        "estimated_cost": prompt_cost["cost"],
+                        "estimated_cost_usd": prompt_cost["cost_usd"],
                         "inference_service": inference_service,
                         "model": model,
                     }
@@ -338,21 +352,21 @@ class Jobs(Base):
             df.groupby(["inference_service", "model"])
             .agg(
                 {
-                    "estimated_cost": "sum",
+                    "estimated_cost_usd": "sum",
                     "estimated_input_tokens": "sum",
                     "estimated_output_tokens": "sum",
                 }
             )
             .reset_index()
         )
-        df["estimated_cost"] = df["estimated_cost"] * iterations
+        df["estimated_cost_usd"] = df["estimated_cost_usd"] * iterations
         df["estimated_input_tokens"] = df["estimated_input_tokens"] * iterations
         df["estimated_output_tokens"] = df["estimated_output_tokens"] * iterations
 
         estimated_costs_by_model = df.to_dict("records")
 
         estimated_total_cost = sum(
-            model["estimated_cost"] for model in estimated_costs_by_model
+            model["estimated_cost_usd"] for model in estimated_costs_by_model
         )
         estimated_total_input_tokens = sum(
             model["estimated_input_tokens"] for model in estimated_costs_by_model
@@ -362,7 +376,7 @@ class Jobs(Base):
         )
 
         output = {
-            "estimated_total_cost": estimated_total_cost,
+            "estimated_total_cost_usd": estimated_total_cost,
             "estimated_total_input_tokens": estimated_total_input_tokens,
             "estimated_total_output_tokens": estimated_total_output_tokens,
             "model_costs": estimated_costs_by_model,

{edsl-0.1.37.dev3 → edsl-0.1.37.dev4}/pyproject.toml

@@ -13,7 +13,7 @@ keywords = [ "LLM", "social science", "surveys", "user research",]
 license = "MIT"
 name = "edsl"
 readme = "README.md"
-version = "0.1.37.dev3"
+version = "0.1.37.dev4"
 [[tool.poetry.packages]]
 include = "edsl"
 

edsl-0.1.37.dev3/edsl/__version__.py

@@ -1 +0,0 @@
-__version__ = "0.1.37.dev3"